Sturla Løkken - Flexible empirical Bayes estimation of local fertility schedules: reducing small area problems and preserving regional variation

  • Presenting author: Sturla Løkken (Research dept. Statistics Norway)

  • Authors: Stefan Leknes & Sturla A Løkken

  • Session: A02D - Spatial - Monday 16:30-18:00 - Erika-Weinzierl Hall

  • Slides: PDF

High-quality local demographic schedules are foundational to many microsimulation models, but small area problems pose a challenge to estimation. Small area estimation methods typically regularize estimates which reduce cross-area variation and may result in severe underestimation of geographic differences. We propose using empirical Bayes methods based on a linear model with three hierarchical geographic levels to predict small area fertility schedules. The model limits sampling variability, captures regional variations effectively, allows for heterogeneity in fertility age patterns, and outperforms alternative models in terms of prediction accuracy. The beneficial properties of the model are demonstrated through simulations and estimations on full-count Norwegian population data.